In most applications, the encoder is mounted entirely between two rings inside a space, whereas, in the solution described above, the encoder extends outside of the space until it is positioned in front of the axial lug: since the protective cover is usually mounted on the outside of the outer ring and is normally arranged in axial abutment against an axial surface in front of the axial lug, the above-described positioning of the encoder prevents not only any axial abutment of the cover against the lug, but also the possibility to optimally control the distance between the cover and the encoder, with the risk of reducing the strength of the signal produced by the rotation of the encoder and picked up by a sensor which faces the encoder and is outside the cover.
Moreover, since it is no longer possible with the above-described configuration of the sensor to mount the cover in axial abutment against the lug of the outer ring, it is necessary to shape the cover differently to that presently known, running the further risk of no longer being able to precisely fit the cover on the lug, with the undesirable result that contaminants may possibly infiltrate between the cover and the outer ring, to the detriment of the insulation of the sensor and of the leaktightness of the unit.